Throttle Pedal [SOLVED]

[EV_Builder]

Any one any idea if the Tesla Throttle / Pedal is usable with the OI?

[Jack Bauer]

Yes that works fine. First one I ever used:)

[EV_Builder]

Ok txs for confirming.

Only thing is that it won't let you measure the resistance so it's not only a simple pot in there...

This is the pin out for reference.

Voltage 0.7v to 3.8v approx.

1= Ref.
2=Sign
3=Return
4=Return
5=Signal
6=Ref.

Note: See the swap of polarity on the second set!

[jon volk]

Hall effect sensor

[vin]

Hi All,

I'm having of bit of trouble with setting up my throttle. I've managed to get everything working in MANUALRUN on FOC 4.94.R, however when I connect my throttle pedal and plot POT and POT2, the pedal activity appears to be registering in the plot graph, however I cant seem to get any activity from the motor in RUN mode. Any ideas please?

Thanks

[jtamborra]

I'm having a weird problem with my throttle pedal while bench testing LDU at 80volts.
Dual hall effect pedal from old maxima, with two pot channels test and plot perfectly.
I configured the dual channel potmode for throttle signal redundancy, calibrated the two pot signals, everything seems fine, running motor, etc...
The problem is, there's a giant dead spot in the middle portion of throttle travel.
I've run a plot showing both throttle pots rising slowly as I depress the pedal VS amps to the motor. there's a deadspot thru the major central part of the travel, and it feels software induced. What am I missing? I did need to reverse the direction of motor rotation, is a side effect of that?
Is it a parameter problem? I've attached the plot and the param file if anyone has any chance to look, I'd be very appreciative.
Thanks-john

Blue=POT Brown=POT2 Red=AMPS

[johu]

The parameters ampmin and slipstart are your friends.
First try lowering slipstart in 5% decrements, maybe down to 30% and see if that fixes your problem. Then you can also try raising ampmin like up to 10%

[jtamborra]

Thanks johu, I've visited Patreon to express appreciation for digging into this issue, and look forward to doing so again in the future.

I tried your suggestion, but not much change.
As a control experiment I reverted to SingleRegen mode, unplugged the throttle POT2 pin, recalibrated the min/max values, and it works perfectly linear like it should.
Plotting DualChannel potnom shows a problem with POTNOM, maybe the pedal I'm using doesn't stay within the required offset tolerance between the two signals , thus becoming invalid unless they align -more closely- at the beginning of pedal travel, and at the end... just not in the middle.
Ive attached plots both singleregen and dualchannel POTNOM, for what they're worth...And I've ordered an E46BWM pedal in hopes of finding the proper pair of signals. Is there a tolerance setting between these signals to accept some slop in the paired signal offset?

SingleRegen POTNOM plot

DualChannel POTNOM plot

[Isaac96]

The offset tolerance is 10 and cannot be set by user as of 4.85. If you'd like to edit the software and compile it yourself, the value is in throttle.cpp in the source code.

Relevant code:

Code: Select all

bool Throttle::CheckDualThrottle(int* potval, int pot2val)
{
   int potnom1, potnom2;
   //2nd input running inverse
   if (potmin[1] > potmax[1])
   {
      potnom2 = 100 - (100 * (pot2val - potmax[1])) / (potmin[1] - potmax[1]);
   }
   else
   {
      potnom2 = (100 * (pot2val - potmin[1])) / (potmax[1] - potmin[1]);
   }
   potnom1 = (100 * (*potval - potmin[0])) / (potmax[0] - potmin[0]);
   int diff = potnom2 - potnom1;
   diff = ABS(diff);

   if (diff > 10)
   {
      *potval = potmin[0];
      return false;
   }
   return true;
}

[jtamborra]

Okay, that's super helpful, thanks!

However, the code comment says "//2nd input running inverse"

My dual hall pedal doesn't read as inverted signals, the 2 signals read as scaled -somewhat- linearly, POT is about 2x POT2
I thought all dual hall effect pedals worked like this, does the BMWE46 have this inverted signal type? (I've ordered one to find out, but I'm curious if anyone knows?)

POT and POT2

Thanks to all.
john

[Isaac96]

That comment refers only to the lines directly below it; those check whether potmin is higher than potmax, meaning it's inverted.

For the purposes of checking the slop between the two values:

Code: Select all

  if (diff > 10)
   {
      *potval = potmin[0];
      return false;
   }

is specifically the relevant area.

[johu]

"10" means 10% difference in slope.
Maybe the problem is that pot1 saturates at 3.3V. Put a 2k7 or 4k7 in series with it and recalibrate. This forms a voltage divider with the on board 10k pull-down resistor.

Thanks for jumping in on Patreon!

[jtamborra]

Thanks for the explanations, I thought about a resistor to more closely match the signals.
So, I wired it up, 2k7 resistor to POT, and first test plotted very good, but now after a reboot i seem to have upset my always (for months) reliable WIFI connection. Not lost, but VERY SLOW data rate, no parameters right away, maybe an update every 5-20 seconds...
Now after reversing my steps I still have very slow parameters with the drive unit, have I upset the DU, or does it sound like a component failure? I've searched, but found no real answer. I've disassembled inverter cover to check the board and connections, reseated, flashed to current firmware, and updated the web files as described in the video with parameter subscriptions.
Up until now, I have been bench testing, occasionally with 80v, occasionally with no HV connection, operating everything fine...
but now the contactors will not complete a precharge sequence... Have I cooked it somehow or locked up the DU with an out of range throttle signal?

Apologies if this has been covered elsewhere, otherwise maybe new unexpected behavior, even if I need to buy a new board, thanks
The v4b board shows led1 solid red, led2 flashing red and the wifi board solid red. I've attached parameters too.

params (2).json

(1.13 KiB) Downloaded 394 times

[johu]

You can't lock yourself out with wrong parameters. Sounds like sketchy power supply really. Thou shall measure voltages: 3V3, 5V, 12V

[jtamborra]

Okay, thanks for that suggestion. I have tried with another good, tested power supply(13.4vdc), with all other connections removed besides 12v to pin1 and GND to pin 22 on 23pin connector. I'm still having the VERY slow parameters update thru the web interface...which is probably why my -previously fine- precharge sequence isn't completing.
I think the LDU control board is fine, but my inverter logic is somehow choked up.
I'm beginning to think I may have shorted something LDU/logic based whilst monkeying with the throttle pedal pins and a resistor..
Can you confirm I should be getting parameters normally/quickly over wifi with only 12v and GND to the LDU?
Maybe a hint where to poke test probes on the clear plastic encased interface module?

[jtamborra]

I am pretty stumped with the behavior from my LDU. So much so, that I recreated the entire harness from scratch along with my new E46 BMW pedal just to cross check my work.
Same result though.
Everything powers up, although precharge - HV sequence fails after 5 seconds, it worked fine for months, and Ive double checked my settings for the 80v ive been working fine with for months.
The wifi/web interface appears, but doesn't populate the parameters nor spot values for like 20 secs or so.
I am able to load previously know good parameters, although it takes about 20 mins, one parameter at a time...
I had been using an old version of firmware with the 'RESTORE DEFAULTS" button, and pressed it, could I have ruined a flashed firmware inside the LDU somewhere?
thanks if anyone can help troubleshoot

[johu]

I removed "restore defaults" for a reason. It loads bogus parameters for the LDU and now you have to apply the parameters from Damien github or the wiki.
Precharge dropping out after 5s means udcsw/2 is not reached after 5s. Either because voltage scaling is off or because udcsw is too high or because no precharge is taking place.

About the wifi issues, maybe there is a wifi transmitter disturbing communication. Have you measured the 3V3 rail? E.g. by measuring between the tab of the 1117 regulator and GND.

[jtamborra]

Success!
Everything is functioning perfectly.
I found an intermittent ground fault in my test bench HV switching hardware which explained my precharge fail symptom.
Also, the chromebook I was using has lost ability to communicate with my v4b board at a useful refresh rate. Different computer works fine. (??)
So, the two problems at once lead me down the "What did I burn" rabbit-hole.
It all works great now.
Thanks for the continued support!
Here's a successful plot of my dual-channel throttle and potnom values I was seeking in the first place.

Dual channel throttle and potnom

[johu]

Great

[m.art.y]

Regarding throttle pedal signal. I tried feeding the throttle pedal +5V and ground from a separate source and I only fed signal wire into accelerator input on gen3 prius inverter but this did not work. So it is necessary to power the pedal from the openinverter board also in order for the throttle pedal to work?

[johu]

Yes, GND must be common.